Premium
Feature Preserving Octree‐Based Hexahedral Meshing
Author(s) -
Gao Xifeng,
Shen Hanxiao,
Panozzo Daniele
Publication year - 2019
Publication title -
computer graphics forum
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.578
H-Index - 120
eISSN - 1467-8659
pISSN - 0167-7055
DOI - 10.1111/cgf.13795
Subject(s) - octree , hexahedron , computer science , injective function , robustness (evolution) , mesh generation , boundary (topology) , feature (linguistics) , surface (topology) , computational science , pipeline (software) , toolchain , algorithm , computer graphics (images) , geometry , software , finite element method , mathematics , engineering , mathematical analysis , structural engineering , linguistics , philosophy , biochemistry , chemistry , pure mathematics , gene , programming language
We propose an octree‐based algorithm to tessellate the interior of a closed surface with hexahedral cells. The generated hexahedral mesh (1) explicitly preserves sharp features of the original input, (2) has a maximal, user‐controlled distance deviation from the input surface, (3) is composed of elements with only positive scaled jacobians (measured by the eight corners of a hex [SEK*07]), and (4) does not have self‐intersections. We attempt to achieve these goals by proposing a novel pipeline to create an initial pure hexahedral mesh from an octree structure, taking advantage of recent developments in the generation of locally injective 3D parametrizations to warp the octree boundary to conform to the input surface. Sharp features in the input are bijectively mapped to poly‐lines in the output and preserved by the deformation, which takes advantage of a scaffold mesh to prevent local and global intersections. The robustness of our technique is experimentally validated by batch processing a large collection of organic and CAD models, without any manual cleanup or parameter tuning. All results including mesh data and statistics in the paper are provided in the additional material. The open‐source implementation will be made available online to foster further research in this direction.